Utilization of steric compression to assign the absolute configuration

Cobalt(III) complex of a new stereospecific hexadentate ligand containing the aspartic acid unit. J. Ivan Legg and John A. Neal. Inorganic Chemistry 1...
0 downloads 0 Views 383KB Size
1097

epimerization of exo- to endo-methyl reactant (2 -/I-+ 3) occurs in the pyrolysis of 2, whereas the reverse process (3 + 2) is readily detectable in the case of 3 and occurs about 60% as fast as the rearrangement of 3 to 9 and 8. These observations are compatible with a change in mechanism for the major portion of the carbon rearrangement, from a largely concerted one in the case of exo-methyl reactant 2 to a diradical pathway in endo compound 3. The principal observable reactions of the diradical intermediate, formed by C-7-C-1 cleavage of 3, are recyclization with retention of configuration at the migration terminus to give cis-exo product 9 and recyclization after rotation about C-6-C-7 to give epimerization product 2. The ratio ki,,(exo)/kinv(endo) is given by the product of two ratios: (i) the overall’ pyrolysis rate ratio (0.8) and (ii) the ratio of the fractions of total product obtained with inversion (91 : 1.4). The minimum value of ki,,(exo)/ki,,(endo) is 54. The present results indicate that the configurationinverting transition state from endo-methyl reactant 3 is sterically too strained t o permit the electronically “allowed” concerted process to occur. They also clearly define as CW the sense of the configuration-inverting motion in the rearrangement of exo-methyl reactant 2. Although we would greet proposed alternatives with interest, we presently consider “orbital symmetry” control of the geometry of the transition state an attractive explanation of these findings. (10) The relative rates of rearrangement of 2 and 3 do not permit 2 to be a significant intermediate in the formation of 9 from 3 at low conversion. (11) Conceivably, 2 may originate in cleavage of the C-6-C-7 bond of 3 and not the C-7-C-1 bond, but it seems unlikely that a-acetoxy activation would be superior to allylic activation. 12 (12) J. A, Berson and E. J. Walsh, Jr., J . Am. Chem. Soc., 90, 4730 (1 968). (13) This investigation was supported in part by National Institutes of Health Predoctoral Fellowship S-Fl-GM-33,023-03 from the National Institute of General Medical Sciences.

Jerome A. Berson, George L. Nelson1* Yale Uniuersity, New Haven, Connecticut 06520 University of Wisconsin, Madison, Wisconsin 53706 Received December 4, 1969

Utilization of Steric Compression to Assign the Absolute Configuration and Ring Conformation of Some Transition Metal Complexes Sir: In recent years in the field of transition metal stereochemistry considerable research has been undertaken to determine the absolute configuration and ring conformation of transition metal complexes. The assignments are generally made by analysis of the ORD or CD spectra of the complex’ or by X-ray determination of its structure. Recently, however, nuclear magnetic resonance (nmr) has been utilized to determine the absolute configuration of the cobalt(II1) complexes with the ligand ethylenediamine-N,N’-di-L-a-propionic acid (LL-EDDP) in terms of the magnetic anisotropy of the C-N bond.3 Nuclear magnetic resonance has also (1) S. F. Mason, Quurr. Reo. (London), 17, 20 (1963). (2) Y. Saito, K . Nakatso, M. Shiro, and H. Kuroya, Bull. Chem. SOC. Jup.. 30, 795 (1957). (3) L.N . Schoenberg, D. W. Cooke, and C. F. Liu, Inorg. Chem., 7, 2386 (1968).

confirmed the absolute configuration of the cobalt(II1) complexes containing I-propylenediamine and ethylenediamine-N,N’-diacetic acid (EDDA), and trends in the chemical shifts of the methyl signals of certain propylenediamine complexes of cobalt(II1) and platinum(1V) have been related to the absolute configuration of these complexes.5 These assignments made use of the different electronic environments of certain protons in the two diastereomers. In coordination compounds containing organic ligands, there are several ways in which the electronic environment about a proton can be influenced by intramolecular interactions. These include anisotropy effects from bonds within the molecules,6 involvement of the metal d-electron system on the ligand substituents,’ and van der Waals interaction or steric compressions.8 Utilizing the last-mentioned effect of steric compression we have examined the absolute configurations and ring conformations of some cobalt(II1) complexes. This effect, which has been observed in certain organic molec~les,~ is produced when a hydrogen atom is forced into proximity of some other atom in the molecule. The proton involved in the compression is found to resonate at a lower field than when the compression is absent. For the complexes under study the distance betwfen the two nuclei involved in the compression is 2-3 A which results in a shift of 0.1-0.3 ppm. A shift of similar magnitude was observed by Terrill and Reilley ‘O in connection with the mms-l,2-cyclohexanediamine-N,N’-tetraacetatocobalt(ll1) cation and was assumed to be in part due to steric compression. We have prepared a series of bis(amino acid) complexes of cobalt(lI1) containing optically active amino acids and find that it is possible to assign the absolute configuration of these complexes using nmr steric compression. The compounds synthesized were trans-(0)[Co(en)(AA),]+ and trans-( 0)-[ Co(en)(AA)(gly)]+ (Table I) which were prepared from the reaction of the optically active amino acid (AA) with either [Co(en)(AA)(H20)CI]+ or [Co(en)gly(HzO)C1]+ and separated from the various cis isomers by ion-exchange chromatography. The compounds were characterized by their elemental analyses and visible absorption spectra which exhibited bands at 360 and 533 my with a shoulder at 450 mp characteristic of trans-Co02Nh complexes. The diastereomers were separated on a cationexchange resin and distinguished by the sign of the dominant component of the CD spectrum in the first spinallowed d-d transition region, The relative arrangement of the a protons in the two diastereomers is seen in Figure 1. In these compounds one would expect that the amino acid ring would not be planar but slightly puckered, placing the substituent on the amino acid in a equatorial position and forcing the a proton into a position adjacent to the amino pro(4) J. I. Legg, D. w . Cooke, and B. E. Douglas, ibid., 6, 700 (1967). (5) J. G. Brushmiller and L. G. Stadtherr, Inorg. Nucl. Chem. Lerr., 3, 525 (1967). (6) J. A. Pople, Proc. Roy. Soc., A239, 541 (1957). (7) D . A. Buckingham and P. J. Stephens, J. Chem. Soc., 4583 (1964). ( 8 ) T. Schaefer, W. F. Reynolds, and Y.Yonemoto, Curl. J . Chem., 41, 2969 (1963). (9) S. Winstein, P. Carter, F. A. Anet, and A. J. Bourn, J . Amer. Chem. Soc., 87, 5247 (1965); W. Nagata, T. Terasawa, and K. Tori, ibid., 86, 3746 (1964). (10) J. B. Terrill and C. N. Reilley, Inorg. Chem., 5, 1988 (1966). (11) N. Matsuoka, J . Hidaka, and Y. Shimura, Bull. Chem. SOC.Jup., 39, 1257 (1966).

Communications to the Editor

1098 0 //

P

0

//

(a)

Figure 1. The diastereomers of bis(L-amino acid)ethylenediaminecobalt(II1): (a) A(Ca); (b) A(Ca). The atoms in bold type are those involved in compressions. The bond lengths to these atoms have been exaggerated to emphasize the interactions involved in compression.

ton on the other amino acid or diamine.12 The presence of steric compression is seen by studying the nmr signals of the CY proton or in some cases the substituent group of the amino acid. In considering the biS(Lalanato) complex, the A(C3) isomer places the a proton Table I. Nuclear Magnetic Resonance Data of Some Cobalt(I1) Amino Acid Complexes

Complexa

trans-(O)-[Co(en)(~-ala)z]+ trans-(0)-[Co(en)(gly)(~-thr)]+ truns-(O)-[Co(en)(glyX~-ser)]+ [Co(en)z(~-ala)]* + [C~(en)~(~-val)] *+

Nmr of complex -with-CDbpc +CD

3.81d 1.53e 3.75d 4.03d 4.15d 1.48'vf 3.78d 0.93e

Ref

3.61 1.55

h

3.63 3.85

h h h

3.98 1.5 3.57

12

1.02

1.13e

[Co(~,~-dimethyltrienX~-ala)]*+ 4.OdJ.o

1.15 3.7

13

L-ala = L-alanine, L-thr = L-threonine, L-ser = L-serine, gly = glycine, L-Val = L-valine, en = ethylenediamine, L,L-dimethyltrien = ~,~-2,9-dimethyltriethylenetetramine.All complexes assigned A(C3) absolute configuration. Shifts in ppm from DSS. aProton signal of amino acid. # Methyl signal. Estimated from published spectra. CD and nmr from compound which is enantiomeric with complex having negative CD. This work,

'

between two amine protons of the diamine at some distance from them while in the A(C3) isomer this proton is near one of the amine protons of the amino acid. The latter isomer should then exhibit a steric compression relative to the former. Similar reasoning can be applied to the methyl group on the L-alanine, although, because of its near-equatorial position, the shifts will not be as great. The steric compression would be greater for the methyl group in the A(C8) isomer as the methyl group is most coincident with the amine proton. The data in Table I show that the bis(L-alanato) isomer which exhibits the lower field signal for the CY proton (at 3.81 ppm) also has the higher field methyl signal (at 1.53 ppm) and can be assigned the A(C3) absolute configuration. The other bigamino acid) complexes are assigned absolute configurations in like fashion. These assignments are consistent with the work of Mason' who assigns the A(Ca) absolute configuration to that (12) H. C. Freeman, M. R. Snow, I. Nitta, and K. Tomita, Acta Cryst., 17, 1463 (1964).

Journal of the American Chemical Society

1 92:4

(b)

Figure 2. (a) A(C3)trans-(0)-[Co(r-pn)(EDDA)]+. (b) A(C3) trans(O)-[Co(d-pn)(EDDA)]+.

isomer which has a negative C D band for the A + E transition in the first spin-allowed d-d transition region. It was of interest to see if this phenomenon could be applied to other systems already reported on in the literature. For this technique to be applicable the compound must exist as two isolated isomers in which one or more groups have different amounts of steric compression which can be estimated through the use of molecular models or similar aids. The group which undergoes the steric compression must also give an nmr signal which is readily assignable. The a proton and methyl substituent on alanine are such groups as is the N-methyl signal in N-substituted compounds. Amine protons should also be applicable assuming that the signal can be definitely assigned. A survey of the literature reveals several cobalt(.III) complexes whose absolute configurations could have been assigned on the basis of steric compression. Buckingham, Durham, and Sargeson13 studied the bis(ethylenediamine)cobalt(III) complexes with L-alanine and with L-valine (Table I). The stereochemistry of the CY protons and methyl groups in the diastereomers is the same as for the bis(amino acid) compounds. These authors relate the absolute configuration of these complexes to increased separation of the methyl signals in the A(C3) [ C ~ ( e n ) ~ ( ~ - v a lcation, ) ] ~ + whereas it is seen that the shift in the a-proton signal due to compression is a much more sensitive signal to consider. Similarly, Asperger and LiuI4 have studied the L-ciscobalt(II1) complexes of ~,~-2,9-dimethyltriethylenetetramine with D- and L-alanine (Table I). Owing to the stereoselective nature of the substituted trien, the a-cis complexes occur only in the A(C3) absolute configuration. The L,L-dimethyl(trien) complex with L-alanine exhibits a steric effect similar to the A(C3)[Co(en),al ala)]^+ cation, while the complex with D-alanine can be compared with the A(C3) [Co(en),(~-ala)]~+ compound. The complexes of EDDA or LL-EDDP(Table 11) with propylenediamine have steric considerations analogous to trans-(O)-[Co(en)(~-ala)~]~+. Figure 2 indicates the arrangement of the EDDA in the A(C3) absolute configuration. For LL-EDDP the methyl group occupies position Hb while in DD-EDDPit occupies position Ha. The diamine in Figure 2a is I-pn while in Figure 2b it is d-pn giving the ring backbones the k and k' conformation, respectively.'5l16 When the diamine is in the k' (13) D. A. Buckingham, L. Durham, and A. M. Sargeson, Aust. J . Chem., 20,257 (1963), as tabulated by D. A. Buckingham, L. D. Marzilli, and A. M. Sargeson, J . Amer. Chem. SOC.,89,708 (1967). (14) R. G. Asperger and C. F. Liu, ibid., 89, 708 (1967).

February 25, 1970

1099

(a)

(b)

Figure 3. Regions of the trans,trans-[Co(Meen)~Xz] + isomers exhibiting steric compression: (a) meso isomer; (b) d,l isomer.

conformation, the steric compression is most pronounced and the signal of the Ha proton appears at a lower field than that in the other diastereomer. If the methyl group occupies position Ha, as in the DD-EDDP compound, the position of the methyl group signal in the complex with the k' conformation is at a lower field than the other diastereomer.

[Co(en)(EDDA)]+ [Co(l-pn)(EDDA)]+ [Co(d-pn)(EDDA)]+d , s \CO(~~)(LL-EDDP)]+ [CO(/-~~)(LL-EDDP)] +

[co(d-pn)(~~-EDDP)l+ [c o ( e n ) ( ~ ~ - E D D p ) ] + [co(l-pn)(~~-EDDp)]+ [CO(&~~)(DD-EDDP)]+

Ha

Hb

4.19 4.26 4.23 4.16 3.99 4.15 4.08 3.93

3.37 3.36

CHs'

Ref 4 4

3.34 1.45 1.45 3.53 3.54 3.51

James C. Dabrowiak, Dean W. Cooke Deparfmenf of Chemistry, Wesfern Michigan University Kalamaroo, Michigan 49001 Received October 2, I949

Transient Bromine Atom Charge-Transfer Complexes Observed by Pulse Radiolysis'

Table II. Nmr Data for trans-(0)-[Co(en)(EDDA)]+-TypeComplexesn,Ir Complex

one nitrogen atom and the proton on the neighboring nitrogen. The nmr data of these compounds show that this methyl group is involved in a compression when compared with the racemate (Table 111). Acknowledgment. The financial support of the National Institute of Health (Grant AM-12262) in this research is gratefully appreciated. The cooperation of Dr. William Kreuger of the Upjohn Company, Kalamazoo, Mich., in allowing the use of a Cary 60 for recording CD measurements is also appreciated.

1.47 1.55 1.61 1.53

4 3 3 3 3 3

Sir: Pulse radiolysis2 has been used to study transient charge-transfer complexes in organic reactions. It is known from kinetic spectroscopy data3-6 that iodine and chlorine atoms can form complexes with various electron donors. Only little is known about bromine atom complexes.' Bromobenzene as well as various liquid systems with aliphatic and aromatic bromine compounds were studied at room temperature. In all

/

Shifts in ppm a Complexes with A(C,) absolute configuration. from DSS. c k' diamine conformation. d k diamine conformation. e Signal from enantiomeric compound. f Methyl signal from EDDP.

The technique is also useful for a series of rrans-bis(diamine)cobalt(III) complexes. Buckingham, Marzilli, and Sargeson" have prepared the meso- and d,ltrans,2rans-[C0(N-Meen)~X~]+ compounds where X is NOz-, C1- or 1/2N02-,1/2C1-. Corey and Bailar'b have discussed the different interactions between the possible conformations for the diamine rings. They concluded that the most stable situation exists when both rings have the same conformation (kk or k'k') as found in the d,Z isomers (Figure 3b). The meso isomer (Figure 3a) has one ring as k and one as k', which should be less stable than the racemate because of the interaction in the meso isomer between the methyl group attached t o

.05

i

/-

Iu 350 400 450 500 550 600 650 Wavelength (nm)

Table 111. Nmr Spectra of trans,trans-[Co(N-Meen)zX,]+ X2 ~~~

mew

d,lc

2.34 2.4 2.57 2.4

2.23 2.29 2.46 2.37

~

N0 2 '/zNOz-, '/*Cl-

c1-

Shift in ppm from DSS. * N-Meen = N-methylethylenediamine, Methyl group signal. Q

(15) E. J. Corey and J. C. Bailar, Jr., J . Amer. Chem. Soc., 81, 2620 (1959). (16) The k designation used in this report is that indicated in Figure 3 of ref 15 which is enantiomeric with that shown in Figure 2 of that reference. ..~

(17) D. A. Buckingham, L. D. Marzilli, and A. M. Sargeson, Znorg. Chem., 7,915 (1968).

Figure 1. Transient absorption spectra at the end of an electron pulse: (1) in pure deaerated bromobenzene; (2) in oxygensaturated bromobenzene.

(1) Work supported in part by the Swiss National Funds. Publication I1 of a series on Pulse Radiolysis of Organic Halogen Compounds, (2) Pulse radiolysis apparatus by J. P. Keene, J. Sci. Insfrum., 41, 493 (1964): 4-MeV LINAC, 2-psec pulse width, 0.4-bsec time resolution; standarddose, 4 X lO17eV/g. (3) R. L. Strong, S. J. Rand, and J. A. Britt, J . Am. Chem. Soc., 82, 5053 (1960). (4) T. A. Gover and G. Porter, Proc. Rop. SOC.(London), A262, 476 (1961). ( 5 ) R. E.Buhler and M. Ebert, Nature, 214, 1220(1967). (6)R. E.Biihler, Helu. Chim. Acta, 51, 1558(1968). (7) J. S . Bartlett, Dissertation, Rensselaer Polytechnic Institute, 1962; cited in R. L. Strong, J. Phys. Chem., 66, 2423 (1962); results confirmed by H.Tsubomura, efal.,J. Amer. Chem. SOC.,91,265 (1969).

Communications to the Editor